270 ALASKAN GLACIER STUDIES up to or possibly just beyond the outer push moraine observed by Gilbert in 1899. Our photograph (PI. CVm, B), which belongs with this remarkable series of four, was therefore necessarily taken from a little farther to the right and higher than the others. It shows, however, the same group of inclined and overturned, dead or dying trees which are in all the other pictures and which were killed by the advance dated by Gilbert as seven years before 1899, or in 1892. In August, 1909, the ice front had just passed this 1892 maximum. On July 4 and September 5, 1910, the advance had gone still farther, as is shown below. The measured record at this point is as follows. Date of Observation Nature of Change Amount Observer 1892 (probably) to July 25-7, 1899 Retreat 3-^iOO ft. Gilbert 1899 (July 25-7) to July 10/ 1905 Retreat 160 ft. Grant 1905 (July 10) to July 15, 1908 Advance 112 ft. Grant 1908 (July 15) to June 24, 1909 Advance 310 ft. Grant and Higgins 1909 (June 24) to Aug. 23, 1909 Advance 70ft. Tarrand Martin 1909 (Aug. 23) to July 4, 1910 Advance 200ft. Martin 1910 (July 4) to Sept. 5, 1910 Advance 50ft. Martin The photographs at these seven periods of observations, show many interesting details concerning the forms of retreating and advancing glacier fronts, the changes of marginal drainage, the accompanying push moraines, the effect on forests and the rapid reoccupation by annual plants of a deglaciated surface. The half mile of ice front of Columbia Glacier on this islet north of Heather Island showed the following conditions on August 22-23,1909, July 4-9,1910, and September 5, 1910. Beginning where the main ice cliff emerged from the water of the bay on the western side, there was a push moraine of beach sand and gravel 15 or 20 feet high in August, 1909, where the photographs show that there was no such moraine in 1899, 1905, and 1908. The ice was in active contact with this. An older push moraine, probably built by the 1892 advance, still extended some distance out toward the water's edge in 1908 but it did not extend to the coast, as did that of 1909, which is farther south. A tiny push moraine within this older one also shows in Grant's 1908 picture, the advance having then apparently commenced. At the time of Grant's J visit in June 1909, the outer moraine had advanced and grown, being about 25 feet high. It was formed between July 15, 1908, and June 24, 1909, and continued to be pushed forward from June 24 to August 23, 1909, Between August 23, 1909, and July 4, 1910, the glacier margin upon the beach continued to advance, shoving the push moraine upon the beach ahead of it and truncating more and more of the forest edge. Measuring from a graywacke rock ledge upon the beach which was about 300 feet from the ice in August, 1909, the glacier margin advanced 135 feet in the ten months up • Grant, U. S., and Higgina, D, F., Glaciers of the Northern Fart of Prince William Sound, Bull. Anjer. Geog. Soc.. Vol. XL13,1910, p. 785.